Measure what leads to a decision.

Separate the result from its drivers

Net revenue retention, or NRR, compares revenue from the same customer base at the beginning and end of a period, including expansion and contraction. It shows whether that existing base is growing, holding steady or shrinking, independently of new customer acquisition.

I use NRR as the top of the hierarchy, then inspect the components underneath it. An expansion result can make the total look healthy while retention weakens. The components make that distinction visible.

• Gross revenue retention: revenue retained from the existing base, excluding expansion.
• Logo retention: the proportion of customers retained over the period.
• Expansion rate: additional revenue from existing customers relative to the revenue at the start of the period.

These measures answer different questions. I want the reporting to preserve those differences so that an encouraging aggregate does not close down the investigation too early.

Choose leading indicators that can change the work

I usually begin with time to first value, adoption and stakeholder engagement. They connect the early customer experience, the depth of product use and the strength of the relationship.

Time to first value

I measure the time from contract close to a meaningful first milestone. The milestone needs a specific definition. Finishing an internal checklist is useful only if it corresponds to progress the customer can recognize.

Product adoption

I look at the depth and breadth of feature use across the customer's users. Core workflow activity is more informative than a login count on its own.

Stakeholder engagement

I look at how many stakeholders are meaningfully engaged and at what level. A relationship that depends on one contact needs different attention from one with an active champion and executive sponsor.

I add other measures when they answer a specific operating question: onboarding completion, business review completion, renewal forecast accuracy and time spent in red health. I avoid adding a measure just because the data is available.

Diagnose from the top; act at the account level

When NRR declines, I first separate retention from expansion. If gross retention is weakening, I inspect account health, onboarding progress and the reasons accounts are leaving. If expansion is weak, I inspect value evidence, stakeholder engagement and the business review process.

Those connections guide an investigation; they do not prove the cause. A metric tells me where to look. The account evidence tells me what action is appropriate.

The action then moves back up the hierarchy. A clearer first-value milestone changes onboarding work. A stronger relationship plan changes engagement. An intervention addresses an account risk. Retention and expansion results show whether those actions are translating into the intended business outcomes.

The health scorecard brings the account signals together. The business review is a place to test the value evidence with the customer.

Treat benchmarks as context

I use benchmark ranges to prompt questions, not to set an account-independent definition of success. Contract length, customer size, market and product all change how a retention figure should be read.

A multi-year contract base has fewer renewal events in a given period than an annual contract base. A business serving smaller customers can have a different logo-retention pattern from one serving enterprise accounts. Comparing the headline number without that context can produce the wrong conclusion.

I establish a baseline from the actual customer base and review the direction over time. The question is whether the underlying retention and growth motion is improving, and whether the leading indicators explain that movement.

Give every measure a place in the cadence

I use the weekly cadence to review changes that need action, and the portfolio and executive reviews to examine the broader pattern. The same definitions should hold across those conversations.

I start with a small set that can be reviewed consistently. The first implementation cycle is a chance to establish baselines, improve the definitions and see whether the selected indicators help explain account outcomes.

A useful report supports a decision: which risk needs attention, what is delaying first value, where engagement is too narrow or what evidence supports expansion. If the report cannot support that discussion, adding more measures is unlikely to help.